Is whole-colon investigation by colonoscopy, computerised tomography colonography or barium enema necessary for all patients with colorectal cancer symptoms, and for which patients would flexible sigmoidoscopy suffice? A retrospective cohort study

Endoscopic whole-colon investigation: colonoscopy

In accordance with the National Institute for Health and Care Excellence (NICE) clinical guidelines, colonoscopy is the reference standard test for establishing a diagnosis of CRC. 3 The advantages of this procedure are that it can visualise the whole colorectum, has high sensitivity for CRC and also permits biopsy and the removal of lesions. 3 Colonoscopy is, however, an invasive procedure that is associated with a small risk of serious complications, including heavy bleeding (1 in 150), bowel perforation (1 in 1500) and, although rarely, death (approximately 1 in 10,000). 5 Colonoscopy can also be an uncomfortable experience for some patients,6 and the intravenous sedation and pain relief generally required for colonoscopy are also associated with cardiovascular and respiratory complications. 7 Moreover, patients who have been sedated are unable to return to work, operate machinery, drive a vehicle or make important decisions for 24 hours after the procedure. 8,9 Older patients or patients with comorbidities are at increased risk of complications and are also less likely to tolerate the purgative full bowel preparation that is required to cleanse the colon prior to colonoscopy. 10,11 The bowel preparation required for colonoscopy is often the part of this procedure that patients find most difficult,12 and failure to complete it results in loss of procedural accuracy. 10,13,14 There is also a small miss rate of colonoscopy for colorectal neoplasia, as demonstrated by tandem colonoscopy studies. 15

Radiological whole-colon investigations

The currently available radiological tests for imaging the large bowel include barium enema and CT colonography. These tests are recommended by NICE for use as alternatives to colonoscopy for the first-line investigation of older patients who are deemed to be at greater risk of complications – particularly those associated with sedation – for those who are unwilling to undergo colonoscopy or for whom a colonoscopy is deemed not possible. 16 In general, however, radiological imaging diagnostics are limited because further endoscopic investigation is required to collect biopsy specimens and/or remove suspicious lesions in the event that these are detected.

Alternative to whole-colon investigation: flexible sigmoidoscopy

Flexible sigmoidoscopy is an endoscopic procedure that is used to examine the distal colon and rectum. Isolated lesions in the proximal colon therefore go undetected with FS, unless a distal lesion is present that warrants a subsequent whole-colon investigation (WCI).

Flexible sigmoidoscopy can offer both clinical and resource use benefits over WCI. FS is a safer, quicker procedure than colonoscopy, and intravenous sedation or pain relief is not typically required, which makes this procedure potentially more appropriate for patients at a higher risk of sedation-related complications, and also means that the patient can generally return to normal activities immediately. 26,27 The lack of need for sedation removes the requirement for (1) recovery time in the endoscopy unit, (2) the patient to be accompanied home and (3) the patient to refrain from driving or operating machinery for 24 hours, as is necessary with sedation. Bowel preparation is more straightforward for FS than for WCI, as it can be achieved with an enema alone (either self-administered or administered by a health-care professional). 26 As with colonoscopy, biopsy tissue and small polyps can be removed during the procedure, although limited bowel preparation generally means that patients with larger lesions are referred for subsequent colonoscopy. Another advantage of FS is that it is a less complicated and less risky procedure than colonoscopy and can be carried out by an appropriately trained nurse specialist. 28

It has been suggested that there is scope within the NHS to reduce the diagnostic burden for patients and endoscopy services by implementing clinical protocols which incorporate the selective use of FS, in place of WCI, for the initial investigation of patients with symptoms suggestive of distal CRC. 29–33 For WCI to be avoided in favour of FS, diagnostic protocols using FS for first-line investigation must be able to demonstrate favourable risk–benefit profiles, in which the benefits of this less invasive procedure are balanced against the risk of a missed diagnosis of proximal cancer. 34–36 The use of such protocols is likely to be most relevant in clinical practice for which the clinical index of suspicion for proximal colon cancer is low, for example when patient and symptom profiles favour a diagnosis of distal CRC. 33,37

The association between anaemia and proximal colon cancer

Anaemia is a condition in which a person has too few red blood cells, or the oxygen-carrying capacity of their red blood cells is diminished, to the extent that physiological needs are not met. 63 Iron deficiency, resulting for example from chronic blood loss, is the most common cause of anaemia and leads to IDA, in which iron stores are depleted and red blood cell production is accordingly compromised. 63,64 Decreased serum ferritin in the presence of low haemoglobin (Hb) and mean corpuscular volume (MCV) is most reliable for the diagnosis of IDA. 64 Blood loss from the gastrointestinal (GI) tract is the most common cause of IDA in postmenopausal women and in adult men, and in approximately 5–10% of cases IDA is caused by colonic neoplasia. 65 Proximal colon cancer in particular is likely to manifest clinically with the consequences of ‘silent’ GI blood loss such as IDA,37,56 which has been reported to be present in up to 75% of patients with proximal colon cancer. 56,60,66,67 Hb levels are generally lower in those with proximal colon cancer than in those with distal cancer,61,68 although there is evidence to suggest that Hb levels in a significant proportion of patients with proximal cancer are higher than locally and nationally defined diagnostic thresholds. 68 Thus, the diagnostic accuracy of anaemia for proximal colon cancer in clinical practice is likely to be influenced by laboratory thresholds68 and by which haematological parameters are used to determine anaemia status. 61 Similarly, the interpretation of clinical studies investigating the diagnostic value of anaemia for CRC by subsite is hampered by variations in the thresholds and haematological parameters used.

Primary objective

The primary objective of the SOCCER study was to investigate the link between patients’ symptoms at presentation and the risk of cancer in the proximal colon to determine whether or not there are particular symptoms or symptom combinations which indicate that a patient could be adequately cared for by a distal colorectum examination (FS) rather than a more extensive WCI. The primary outcome of the SOCCER study was the diagnostic yield of CRC (proximal/distal) within 3 years of presentation at clinic, by symptoms at presentation.

Secondary objectives

The secondary objectives of the SOCCER study were to:

• measure the prevalence of proximal and distal CRC in referred patients presenting with symptoms suggestive of CRC

• determine the number needed to be examined to diagnose one distal cancer, by symptoms at presentation

• determine the number needed to be examined to diagnose one proximal cancer, by symptoms at presentation

• determine the miss rate of CRC after FS

• determine hypothetical proximal CRC miss rates if only patients with certain symptoms or combinations of symptoms are sent for WCI.

Ethics approval and permission to use patient data without consent

Ethics approval for the SIGGAR randomised controlled trials was obtained from the Northern and Yorkshire Multi-Centre Research Ethics Committee on 15 January 2004 and, subsequently, from individual participating centres. Research ethics approval for the SOCCER study was granted as an extension to the SIGGAR randomised controlled trials by the North East (York) National Research Ethics Service. The SOCCER study was also granted Section 251 support under the National Health Service Act 200673 for the processing of patient identifiable information without consent [references ECC 5–04(E) 2011 and 14/CAG/1043]. To comply with the conditions of Section 251 support, the Cancer Screening and Prevention Research Group at Imperial College London (responsible for all aspects of trial and data management for this study) assessed its data handling procedures against Department of Health information governance standards. The Cancer Screening and Prevention Research Group holds an Information Governance toolkit to demonstrate compliance with these standards. 74

Selection of participating hospitals

Patients were recruited to the SIGGAR trials from hospital trusts in which a radiologist member of SIGGAR had expressed a prior interest in participating. Centres were expected to have an established and efficient fast-track referral system for patients with suspected CRC (usually an identified diagnostic clinic) to facilitate recruitment, and a named colorectal nurse specialist or researcher who would take responsibility for recruitment.

The final 21 NHS hospitals were selected via a ‘sham randomisation’ that identified centres likely to achieve a minimum monthly recruitment target (at least 18 patients). 23 These 21 hospital centres included teaching and general hospitals and were distributed across England (see Appendix 1).

SOCCER eligibility criteria

Patients who were considered potentially eligible for the SIGGAR trials were considered eligible for the SOCCER study, irrespective of whether or not they were randomised, unless they met the SOCCER study exclusion criteria.

Baseline characteristics

Patient baseline characteristics were collected when patients were originally assessed for eligibility for the SIGGAR trials, and data were collected for both randomised and non-randomised patients. This information had been recorded on the bespoke SIGGAR trials pro forma and included patient age, sex, date of referral, the urgency of the referral (‘2-week wait’, ‘urgent’, ‘soon’ or ‘routine’), the referral route, the diagnostic investigations requested, the outpatient clinic type (if applicable), other relevant diagnoses and whether or not the patient had initially been investigated by FS. For randomised patients, details of the main SIGGAR trial interventions (barium enema, CT colonography and colonoscopy) and outpatient appointments were also recorded on the trial pro forma.

Symptoms and clinical signs

Clinical features at presentation were recorded for potentially eligible patients at baseline during eligibility assessment for the SIGGAR trials. The SIGGAR trials pro forma contained tick boxes to record symptoms and clinical signs under ‘details/reason for referral’. Tick boxes were included for ‘rectal bleeding’, ‘abdominal pain’, ‘anaemia’, ‘weight loss’, ‘CIBH’ and ‘positive FOBt’. A free-text field to record additional symptoms was also included on the pro forma. Entries in the free-text field were manually coded by the trial team for use in the analysis. They were categorised into ‘abdominal mass’, ‘bloating/flatulence’, ‘tiredness/weakness’, ‘anal symptoms’, ‘nausea/vomiting’, ‘back pain’, ‘upper GI symptoms’, ‘rectal mass’, ‘family history’, ‘history of polyps’, ‘presence of cancer antibodies’, ‘elevated C-reactive protein’ and ‘liver problems’. A second free-text field to record the details of the CIBH was also included on the pro forma and was manually coded and categorised to ‘looser and/or more frequent’, ‘harder and/or less frequent’, ‘variable’ or ‘unspecified’.

Data pertaining to clinical features at presentation were also sourced from hospital records. Radiology, endoscopy and pathology records were requested for patients in the SOCCER study cohort and were interrogated for information concerning symptoms/clinical signs (specifically abdominal mass, rectal bleeding, abdominal pain, weight loss, a CIBH and rectal mass). Relevant data were extracted from text fields. For further details see Data extraction.

Flexible sigmoidoscopy

Details of FS procedures performed at the time of referral had been recorded on a separate pro forma during the SIGGAR trials and included room entry and exit times; procedure start and stop times; overall assessment of the examination by the endoscopist (‘very easy’, ‘quite easy’, ‘quite difficult’ or ‘very difficult’); assessment of bowel preparation quality by the endoscopist (‘excellent’, ‘good’, ‘adequate’ or ‘poor’); the segment of the colon reached and reasons (if any) the examination could not be completed; overall findings and details of polyps, cancers or biopsies and diverticula (with a severity rating of ‘none’, ‘mild’, ‘moderate’ or ‘severe’); and adverse events occurring during the procedure. Unfortunately, during scrutiny of these records, it was discovered that in many cases the information included had been taken from the electronic endoscopy record and that many items were missing.

Manually collected data

Data were collected manually in the following scenarios.

• The hospital did not have the facilities or specialists to bulk extract the data for us.

• The quoted cost for bulk extracting the data obtained from the suppliers of the system was excessive, making manual data collection more cost-effective.

• It was possible to bulk extract the data only from a data warehouse/reporting system (not the main databases in which the raw data were held) and our findings showed that the data warehouse was not always up to date. In this scenario we collected the data manually from the applications that were linked to the main databases.

• The hospital was unable to find specialists to help with bulk extraction within our required time frame, so we manually collected the data in order to meet our data collection deadlines.

• Some hospitals were able to extract the type of test/examination and date but not provide a report. We used this information to identify the records of interest and narrowed down the task of manual data collection to the selected records.

• The data were held on legacy systems and the hospital did not have a maintenance contract with the suppliers, with the result that there was no option but to extract the data manually.

Study researchers visited hospitals to manually collect data in a bespoke Microsoft Access^® database (2010, Microsoft Corporation, Redmond, WA, USA) or spreadsheet which included patient study numbers. Patient identifiers from the SIGGAR trials were held at hospitals and were used to search for patients on hospital databases. De-identified data were returned to the study team, and the study programmer cleaned and uploaded it to a master SOCCER Oracle database (Oracle Database 11g Enterprise Edition, Oracle Corporation, Redwood Shores, CA, USA).

Sample size

Our original sample size assumed that we would have a total cohort of 8484 patients, in whom 421 distal cancers and 68 proximal cancers would be diagnosed. The analysis plan presented estimates for the precision for the estimated sensitivity under specific regimens, with the precision being conditional on the number of cancers diagnosed. We assumed that under a regimen offering WCI to patients with IDA and/or an abdominal mass we would detect 470 of the total 489 cancers, giving a sensitivity estimate of 96.1% with a 95% confidence interval (CI) of 94.0% to 97.6%. Although the final analysed cohort of 7380 patients was smaller than the proposed sample size, the number of cancers diagnosed was greater than expected, with a total of 429 distal cancers and 127 proximal cancers, thus providing a greater level of precision than originally estimated.

Primary outcome

The primary outcome was the diagnostic yield of distal or proximal cancer within 3 years of presentation at clinic, by symptom category at presentation. 18,19,23 CRC diagnoses were sourced from the HSCIC and from patient medical records. For cancers confirmed by a hospital pathology report but without corresponding verification by HSCIC, the local pathology report was taken as conclusive evidence of cancer. For the purposes of this study, CRCs included all cancers with International Classification of Diseases and Related Health Problems, Tenth Edition,75 site codes C18–C21 and with an International Classification of Diseases for Oncology, Third Edition,76 morphology code of 8000/3, 8010/3, 8070/3, 8123/3, 8140/2, 8140/3, 8144/3, 8210/3, 8261/2, 8261/3, 8263/2, 8263/3, 8480/3, 8481/3, 8490/3, 8510/3 or 8560/3. CRCs were classified as ‘distal’ if they were located in the anus, rectum, sigmoid colon or descending colon. Cancers located proximal to the descending colon were classed as ‘proximal’. Synchronous distal and proximal CRCs were included as separate cancers in the analysis.

Secondary outcomes

The secondary outcomes were the sensitivity of symptoms and symptom categories for distal and proximal cancer, the percentage of patients with cancer who had distal CRC by symptom and symptom category, the number needed to be examined to diagnose one distal or proximal cancer by symptom and symptom category at presentation, the miss rate for CRC at FS in the subgroup of patients with FS performed at baseline and the prevalence of proximal and distal CRC in the study cohort.

Analysis

Outcomes were first analysed separately in the cohort with blood test data and the cohort without blood test data. The findings in the two cohorts were then compared and outcomes analysed in the total combined cohort.

Sensitivity was calculated as the proportion of CRCs by cancer site (proximal/distal) that were identified by a particular symptom or symptom combination. Specificity was defined as the proportion of patients without CRC by cancer site who presented without a particular symptom/symptom combination.

Diagnostic yields were presented as percentages. The number needed to be examined was calculated as the inverse of the diagnostic yield. Binomial exact 95% CIs were calculated for key outcomes. The distributions of categorical variables (patient characteristics, referral details, symptoms, signs, indications and cancer outcomes) were compared between cohorts using Pearson’s chi-squared test or Fisher’s exact test, as appropriate, and all tests were two-tailed. Comparisons were made between: cohorts with and without blood test data; men and women; patients with distal cancer and patients with proximal cancer; and patients with and without FS performed at the time of referral. Data were analysed using Stata version 13.1 (StataCorp LP, College Station, TX, USA).

Patient demographics and referral details

Overall, in the full cohort of 7380 patients, there were more women (59.0%) than men (Table 1). The majority of patients were referred via a colorectal outpatient clinic (84.5%). Just over half of all patients (n = 3976) were referred via the suspected cancer 2-week wait pathway, and a further 18% were considered ‘urgent’ referrals (n = 1315). Only 12% (n = 915) of patients in the full cohort were routine referrals.

Comparing the patients with and without blood test data available, those with blood test data were slightly older, less likely to be referred via a colorectal outpatient clinic and more likely to be referred via a gastroenterology outpatient clinic than patients without blood test data (all p < 0.001). Slightly more patients in the cohort with blood test data were referred via the 2-week-wait pathway than in the cohort without blood test data (p < 0.001).

Colorectal cancer symptoms and clinical signs at presentation

Overall, the most common symptom or clinical sign in the full cohort (n = 7380) was a CIBH: this symptom was reported in > 70% (n = 5382) of all patients (Table 2). Other common features included rectal bleeding (n = 2773), abdominal pain (n = 2126) and weight loss (n = 1148). Overall, the symptom/clinical sign profiles of the cohorts with and without blood test data were similar, although there were some differences. Notably, rectal bleeding and rectal mass were more common in the cohort without blood test results, but weight loss and tiredness/weakness were more common in the cohort with blood test results.

Colorectal cancer diagnoses by anatomical subsite

The prevalence of CRC in the whole cohort was 7.5% (Table 3). Overall, distal cancer was diagnosed in 5.8% of patients and proximal cancer was diagnosed in 1.7% of patients. Distal cancer was less common in the cohort with blood test data than in the cohort without (p < 0.001), whereas proximal cancer was more common in the cohort with blood test data (p = 0.007).

Anaemia and iron deficiency anaemia

Of the 4741 patients for whom blood test data at presentation were available, serum ferritin results were available for 1157 (approximately 24%) (Table 4). Among patients for whom ferritin data were available, low levels of serum ferritin (< 20 µg/l) were reported in approximately 31% (n = 353). MCVs were low (< 80 fl) in 9.0% (n = 176) of men and 8.2% (n = 229) of women. In total, 31.8% (n = 256) of men and 36.5% (n = 311) of women with low Hb levels (< 13 g/dl for men and < 12 g/dl for women) had either a low serum ferritin or a low MCV level. Only 1.1% (n = 13) of men and 2.7% (n = 53) of women with normal Hb also had low serum ferritin or MCV.

Colorectal cancer symptoms and clinical signs by sex and age

Clinical features that were more common in women than in men were CIBH (all subtypes with the exception of harder stools and/or less frequent defecation) (p < 0.001), abdominal pain (p < 0.001) and abdominal mass (p = 0.029) (Table 7). Rectal bleeding was more common in men than in women (p < 0.001). Similarly, more men than women presented with anaemia (p < 0.001) and IDA by the broad definition (p = 0.037).

The proportions of men and women with rectal bleeding or abdominal pain decreased with increasing age (Table 8). Anaemia was substantially more common in older age groups in both men and women; 73% of men (n = 66) and more than half of women (n = 83) aged ≥ 85 years were anaemic (broad definition), compared with 24% of men (n = 131) and 19% of women (n = 161) aged 55–64 years. Similar trends with age for men and women were observed for all other definitions of anaemia and IDA.

The proportions of patients presenting with a CIBH (all subtypes combined) were also influenced by age; however, the trends observed for men and women were not the same. A CIBH was more commonly reported in men aged ≥ 85 years (82.2%) than in men aged 55–64 years (69.3%) but was less common in older women (69% in those aged ≥ 85 years vs. 78.3% in those aged 55–64 years). The trend in men was largely a result of the increase in the proportion of men with harder stools and/or less frequent defecation with increasing age (8.9% in those aged 55–64 years vs. 25.6% in those aged ≥ 85 years), whereas fewer women reported a CIBH to looser and/or more frequent stools with increasing age (44.2% in those aged 55–64 years vs. 29.0% in those aged ≥ 85 years).

There was evidence of some association between age, gender and the presence of other less common clinical features, although in many cases the numbers available for analysis by age group were low. In women, but not in men, the presence of an abdominal mass as a reported clinical sign increased with increasing age. An abdominal mass was reported by 11.0% of women (n = 16) aged ≥ 85 years, compared with 2.1% of women (n = 18) aged 55–64 years. Similarly, there was evidence to suggest that a rectal mass was more commonly reported in older women (7.6% in women aged ≥ 85 years vs. 1.6% in those aged 55–64 years).

Colorectal cancer symptoms and clinical signs in isolation and in combination

Approximately 95% of patients (n = 4486) presented with between one and three features that are referral criteria in the NICE 2015 guidelines76 for suspected CRC; just over one-third (n = 1626) presented with a single NICE criteria symptom/sign (Table 9).

The NICE qualifying features most commonly reported as single symptoms were a CIBH and rectal bleeding. For NICE criteria signs, between 36% and 43% of patients (depending on the definition used) presenting with IDA did so in the absence of any other NICE criteria. The proportions of patients presenting with anaemia/IDA in the absence of other NICE criteria symptoms/signs increased as the definition used for anaemia became stricter through inclusion of a requirement for iron deficiency.

The majority (67.4%) of the 3472 patients with a CIBH presented with this symptom in combination with one or two other NICE qualifying features. Similarly, the majority of patients (73.2%) with rectal bleeding presented with this symptom in combination with one or two other NICE qualifying features. Few patients with weight loss (1.2%) or abdominal pain (7.9%) presented without additional NICE qualifying features. Very few patients (n = 16) with features that were non-NICE qualifying presented without additional NICE criteria.

Patients with NICE qualifying features commonly presented with symptoms in combination with a CIBH, which probably reflected the high prevalence of this symptom in the cohort (73% overall prevalence) (Table 10). For instance, > 60% of patients (n = 1022) with rectal bleeding presented with a CIBH.

Patients with anaemia/IDA commonly presented with these signs in combination with a CIBH, rectal bleeding and weight loss. Sixty per cent (n = 989) of patients with anaemia (broad definition) also had a CIBH, compared with 37% (n = 119) of patients with IDA (strict definition) (see Table 10).

Patient demographics and referral details in those with colorectal cancer diagnoses

Distal cancers were more likely to be diagnosed in men and proximal cancers were more likely to be diagnosed in women (p = 0.013) (Table 11). There were no major differences in subsite diagnoses by referral route (p = 0.21) or urgency of referral (p = 0.62).

Prevalence of distal colorectal and proximal colon cancers by age and sex

In stratified analyses, there were apparent trends in the prevalence of distal and proximal cancer by age and sex (Table 12). In women, the prevalence of distal cancer increased with age; 5.5% of women (n = 8) aged ≥ 85 years were diagnosed with distal cancer, compared with 2.2% of women (n = 19) aged 55–64 years. This trend in distal cancer prevalence by age was less apparent in men; however, distal cancer was more prevalent in men aged ≥ 65 years than in those aged 55–64 years. The prevalence of proximal cancer increased with increasing age in both men and women.

Symptomatic presentation of proximal colon cancer and distal colorectal cancer diagnoses by clinical features in isolation and in combination

A small proportion (9.3%; n = 16) of the 172 distal cancers diagnosed in patients with a CIBH were diagnosed in those with a CIBH alone (Table 14), in comparison to a high proportion (59.9%; n = 103) diagnosed among individuals who also reported rectal bleeding. Of the 52 proximal cancers diagnosed in patients with a CIBH, only four (7.7%) were diagnosed in those with a CIBH alone and 11 (21.2%) in those who also had rectal bleeding. However, when anaemia was also reported, 35 (67.3%) of these proximal cancers were captured.

Abdominal pain and weight loss as single clinical features were not predictive of either distal or proximal cancer (see Table 14). The majority of distal cancers in patients with abdominal pain (78.4%) and weight loss (74.0%) were diagnosed in those who also had a CIBH. Furthermore, the majority of proximal cancers in those with abdominal pain (n = 22) were in patients who also had a CIBH (66.7%) or anaemia (66.7%). Similarly, the majority of proximal cancers in those with weight loss occurred in patients who had this symptom in combination with a CIBH (66.7%) or anaemia (87.5%). No cancers were diagnosed in patients with weight loss alone.

Anaemia was notable out of all the NICE qualifying features for indicating a high ratio of proximal to distal cancers, which was in contrast to any other NICE guideline symptom (see Table 14). The ratios of proximal to distal cancer increased as the definitions for anaemia/IDA were tightened. The highest ratio was observed in patients with IDA by the strict definition (20 proximal cancers vs. 12 distal cancers). By contrast, anaemia by the broadest definition was not as specific for proximal cancer (77 proximal cancers vs. 106 distal cancers). Out of the 77 proximal cancers found in patients with anaemia (broad definition), 22 (28.6%) were found in patients with no other NICE qualifying features. Similarly, 8 out of 20 (40.0%) proximal cancers in patients with IDA were diagnosed in those who had this sign alone. Relative yields of distal cancers were higher in patients with anaemia/IDA (by any definition used) in combination with rectal bleeding. Using the broad definition of anaemia, 59 distal cancers were diagnosed in patients who also reported rectal bleeding, compared with 14 proximal cancers.

The NICE qualifying symptom associated with the highest proportionate number of patients diagnosed with distal cancer versus proximal cancer, regardless of other presenting clinical features, was rectal bleeding: when patients presented with this symptom, 89.5% of those diagnosed with cancer had distal cancer (Table 15).

Rectal bleeding was examined in further detail as a symptom alone and in combination with other symptoms (Table 16). Among patients presenting with rectal bleeding alone, 89.5% (n = 17) of those diagnosed with cancer had a distal tumour. When rectal bleeding and a CIBH were reported together in the absence of anaemia, 95.8% (n = 69) of patients diagnosed with cancer had a distal cancer. This increased to 100.0% (n = 10) when the CIBH was reported as harder and/or less frequent defecation.

A more detailed examination of CIBH revealed generally lower proportions of distal cancer than when presenting symptoms included rectal bleeding (Table 17). When patients presented with a CIBH alone, 83.3% (n = 15) of those diagnosed with cancer had distal cancer. Of patients with cancer who presented with a CIBH and anaemia, between 57.9% and 71.7% had distal cancers, with proportionately more distal cancers in those reporting a change to looser and/or more frequent stools and anaemia (n = 38) than in those with a change to harder stools and/or less frequent defecation and anaemia (n = 11).

It has previously been reported that proportionately fewer proximal cancers are diagnosed in patients without anaemia and/or an abdominal mass. 31,33 Here, of the 187 patients diagnosed with cancer who presented with anaemia or an abdominal mass, 59.4% (n = 111) had distal cancer (Table 18). In patients diagnosed with cancer who had presented without anaemia or an abdominal mass, the proportions with distal cancer varied by symptom combination. Overall, 88.4% of patients (n = 129) without anaemia or an abdominal mass who had cancer had distal cancer; among these patients, a report of rectal bleeding increased this proportion to 96.0% (n = 95). By contrast, in patients without anaemia or abdominal mass who reported a CIBH but without rectal bleeding, only 71.1% (n = 32) of patients with cancer had distal cancer.

Diagnostic yields of clinical features for proximal and distal colorectal cancers

The overall diagnostic yield for distal cancer was more than twice that for proximal cancer (5.1% vs. 2.0%) and, as a corollary, fewer patients needed to be examined to diagnose one distal (20 patients, 95% CI 18 to 23 patients) than one proximal cancer (50 patients, 95% CI 41 to 61 patients) (Table 19). The diagnostic yields were higher for distal cancer than for proximal cancer for all NICE qualifying symptoms. The yields for proximal cancer were < 3% for all NICE qualifying symptoms. The highest diagnostic yield for distal cancer was observed in patients with rectal bleeding; out of the 1660 patients who presented with rectal bleeding, 154 (9.3%) were diagnosed with distal cancer, resulting in a low number who needed to be examined to diagnose one cancer (11, 95% CI 10 to 13). By contrast, only 1.2% patients (n = 20) with rectal bleeding were diagnosed with proximal cancer.

Using NICE qualifying clinical signs, the highest diagnostic yield for proximal cancer was observed in patients with IDA (by the broad definition), of whom 8.5% were diagnosed with proximal cancer. The diagnostic yields in patients who presented with an abdominal mass were comparatively high for both proximal (5.7%) and distal cancer (7.1%). None of the 81 patients with rectal mass were diagnosed with proximal cancer, but this sign was highly predictive of distal cancer (diagnostic yield 18.5%).

In the 1660 patients with rectal bleeding for whom blood test data were available, diagnostic yields for distal cancer were as high, at 34.5% among those with a rectal mass and 11.9% for a CIBH to looser and/or more frequent stools (Table 20). By contrast, diagnostic yields for proximal cancer were approximately ≤ 1%, except in patients with anaemia (2.8%) or an abdominal mass (5.9%), and there was a large degree of uncertainty in the numbers needed to be examined to diagnose one proximal cancer. The diagnostic yield for distal cancer was notably higher in non-anaemic patients who had rectal bleeding in combination with rectal mass (34.5%); however, > 1 in 10 patients with rectal bleeding and anaemia were also diagnosed with distal cancer (diagnostic yield 11.8%).

In patients with a CIBH, the highest diagnostic yields were once again observed for distal cancer (Table 21). Yields of up to 21.4% were observed for distal cancer in patients with a CIBH in combination with rectal mass, although only 15 out of 833 patients who presented with a CIBH alone were diagnosed with distal cancer (diagnostic yield 1.8%). Yields for proximal cancer were generally ≤ 4% in patients with a CIBH.

Of the NICE qualifying features, the presence of anaemia and/or an abdominal mass was associated with the highest diagnostic yield for proximal cancer, with 4.5% of these patients (n = 78) being diagnosed with proximal cancer (Table 22). In patients without anaemia or an abdominal mass, no other NICE qualifying features, either in isolation or in combination, exhibited diagnostic yields for proximal cancer of more than approximately 1%, with the exception of a CIBH to harder stools and/or less frequent defecation (diagnostic yield 2.2%). There was evidence to suggest that NICE qualifying features were more predictive of distal cancer than of proximal cancer in patients without anaemia or an abdominal mass (overall diagnostic yield of 4.3% vs. 0.6% for distal and proximal cancer, respectively). In particular, rectal bleeding, either alone or in combination with other clinical features, was associated with diagnostic yields of up to 17.5% (range 1.7–17.5%) for distal cancer but < 0.8% for proximal cancer.

We examined the proportions of patients diagnosed with proximal and distal cancers by different definitions of anaemia (anaemia broad, anaemia strict and IDA broad) (Table 23). Of the 95 patients with cancer who presented with IDA and/or an abdominal mass, 47.4% (n = 45) were diagnosed with distal cancer. Similar observations were made for patients with strict-definition anaemia and/or an abdominal mass (n = 59; 48.8% distal cancer). There was some evidence to suggest that a slightly greater proportion of cancers diagnosed in patients with broad definition anaemia and/or an abdominal mass were distal (n = 111, 59.4%). Similar proportions of cancers diagnosed were distal in patients with a symptom profile including rectal bleeding irrespective of the anaemia definition used (see Table 23).

Comparing the diagnostic yields and numbers needed to be examined in patients with anaemia or an abdominal mass, with anaemia status defined by the strict definition (Table 24) and IDA by the broad definition (Table 25), diagnostic yields for proximal cancer were similar (6.4% and 7.4%, respectively); both of these estimates are higher than the 4.5% yield when the broad definition of anaemia is used (see Table 22). The diagnostic yields for proximal cancer in patients without anaemia or an abdominal mass by any definition were similar (range 0.6–1.1%). The number needed to be examined to diagnose one proximal cancer in patients without IDA or an abdominal mass was 91 (95% CI 68 to 124; see Table 25), compared with 114 (95% CI 82 to 166; see Table 24) for the strict definition of anaemia and 168 (95% CI 106 to 282; see Table 22) for the broad definition of anaemia.

We also examined the diagnostic yields and numbers needed to be examined for distal and proximal cancers in patients with anaemia (all definitions) separate from an abdominal mass (Table 26). For distal cancers, the highest diagnostic yields were observed in patients with anaemia with/without additional symptoms (range 3.8–6.4%). Anaemia by the broadest definition was most predictive of distal cancer in this group of patients, whereas IDA strict definition was least predictive. The diagnostic yield for proximal cancer for all patients with anaemia, by any definition and regardless of whether or not they had additional symptoms, ranged from 4.6% to 8.5%. For patients with anaemia, IDA by the broadest definition and IDA by the strictest definition were the most and least predictive of proximal cancer, respectively.

Patient profiles of proximal cancers diagnosed in those presenting with rectal bleeding in the absence of anaemia or an abdominal mass

A total of 14 (45.2%) out of the 31 proximal cancers in patients without anaemia or an abdominal mass were diagnosed in those with rectal bleeding (Table 31). Five of these patients had pathology or a clinical finding in the distal colorectum that would probably have warranted a follow-on WCI after an initial FS examination, of whom four were female and four were aged ≥ 69 years. Three of these five patients had cancer located in the caecum and the majority (n = 3) presented with a CIBH to looser stools and/or more frequent defecation in addition to rectal bleeding.

The majority (57.1%, n = 8) of the 14 patients with proximal cancer who presented with rectal bleeding without anaemia or an abdominal mass did not have pathology or a clinical finding in the distal colorectum that would have warranted follow-up by WCI. Three-quarters (n = 6) of these patients were male and all were aged ≥ 63 years. Four of the proximal tumours in this group of patients were located in the caecum and three-quarters (n = 6) of these patients had a CIBH to looser stools and/or more frequent defecation in addition to rectal bleeding; only two patients in this group presented with rectal bleeding alone.

Flexible sigmoidoscopy miss rates for distal cancers

Of the 112 patients diagnosed with distal cancer who had been examined by FS, most (90.2%) had cancer diagnosed by FS (Table 37). Of the 11 distal cancers (9.8%) not diagnosed at FS, eight might have been found if current practice was applied, as patients in whom FS was incomplete or in whom a lesion was detected would have gone on to receive a repeat examination and/or a WCI. Only in the three patients with a ‘complete and normal’ FS examination would cancer have been missed by FS (miss rate 2.7%).

There were 31 patients who were initially examined by FS and subsequently diagnosed with proximal cancer. Two-thirds of these (n = 22) were diagnosed after a ‘complete and normal’ FS examination and would not have gone on to have a WCI after this outcome, unless other diagnostic indicators were present.

Flexible sigmoidoscopy miss rates for proximal and distal colorectal cancer by clinical features

Approximately one in four patients (n = 359) who underwent FS at baseline presented with anaemia and/or an abdominal mass, of whom slightly more were diagnosed with proximal than with distal cancer (n = 25 vs. 23; Table 38). Of the 23 patients diagnosed with distal cancer, 18 (78.3%) were diagnosed in those in whom cancer had been suspected at FS and four (17.4%) were diagnosed in patients who had an incomplete examination. Only one patient with distal cancer was diagnosed after a ‘complete and normal’ FS examination (potential miss rate 4%). Notably, 19 out of 25 patients (76.0%) with proximal cancer had a ‘normal and complete’ FS examination.

Half of the patients diagnosed with cancer (53.5%, 76/142) who were examined by FS presented with the symptom profile of rectal bleeding without anaemia or an abdominal mass (see Table 38). In this group, 73 (96.0%) had distal cancer and the large majority of these cancers (n = 69) were diagnosed at FS. Only three patients with this symptom combination were diagnosed with proximal cancer after a FS examination, and two of these patients had findings at FS that might have warranted a WCI.

In total, 39.0% (n = 578) of patients examined by FS presented with a CIBH (of any subtype) without rectal bleeding, anaemia or an abdominal mass, 18 of whom were diagnosed with cancer. The majority (83.3%, n = 15) had distal cancer diagnosed. Only three patients overall were diagnosed with proximal cancer in this group, all of whom had a CIBH to subtypes other than looser and/or more frequent stools. One patient with a CIBH was diagnosed with proximal and distal cancer after FS examination, although this patient also had a large lesion detected at FS, which might have warranted a WCI; the remaining two proximal cancers were diagnosed after complete and normal FS examinations. Fewer patients who presented with a CIBH without other common NICE referral criteria symptoms/clinical signs (anaemia, an abdominal mass or rectal bleeding) were diagnosed with cancer (see Table 38).

Approximately 3% (n = 44) of patients examined by FS presented with abdominal pain/weight loss (without anaemia, an abdominal mass, rectal bleeding or a CIBH) (see Table 38). No patients with this symptom profile were diagnosed with cancer.